Multi-Branch Fractional Rate Change Filter for High-Speed DSP
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital signal processing technologies face challenges in efficiently converting sampling rates due to hardware limitations, particularly when dealing with high-speed sample data rates, as direct implementation of fractional rate change filters exceeds hardware processing capabilities, necessitating complex resource management and parallel branching.
Innovation Solution
A multi-branch finite impulse response filter is employed to convert sampling rates by inputting the signal into multiple filter branches with different phase coefficients, delaying taps to account for pipelining and phase delays, and storing filtered substreams in memory to generate an output stream with a lower sampling rate, using a rate change filter with multiple branches to manage hardware resources effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If direct implementation of fractional rate change filter is used, then sampling rate conversion accuracy is improved, but hardware processing speed capability deteriorates
Solution Approach 1:
The filter is divided into multiple parallel branches, each processing a subset of the input samples. This segmentation allows the overall filtering operation to be distributed across multiple lower-speed processing paths, maintaining conversion accuracy while reducing the speed burden on individual hardware components.
Solution Approach 2:
The patent introduces a temporal dimension through buffering and interleaving operations. By storing filtered samples in buffers and retrieving them in the correct sequence, the system achieves accurate rate conversion without requiring all processing to occur at the highest instantaneous sample rate.
2Speed
If parallel branching of data paths is used, then hardware processing speed capability is improved, but device complexity worsens
Solution Approach 1:
Multiple filter branches share common resources including the input buffer, coefficient storage, and output buffer. This multi-functionality reduces the overall resource requirements compared to implementing separate complete filter chains, lowering complexity while maintaining parallel processing capability.
Solution Approach 2:
The patent uses simplified copies of the filtering logic across multiple branches rather than full duplications. Each branch contains a reduced set of filter coefficients and processing elements, creating lightweight copies that collectively achieve the desired rate conversion without excessive complexity.
3Measurement precision
If up-sampling is performed first, then sampling rate conversion accuracy is improved, but productivity worsens
Solution Approach 1:
Instead of the conventional approach of up-sampling first then down-sampling, the patent inverts the order by performing down-sampling first through selective filtering of input samples. This inversion reduces the number of samples that need to be processed at high rates, improving productivity while maintaining conversion accuracy through proper filter design.
Data Source
AI summary
A method of filtering an input sample stream having a downsampling rate is disclosed to generate an output sample stream having an upsampling rate that is less than the downsampling rate. The input sample stream is input to a rate change filter having multiple filter branches. The input sample stream is filtered at each of the multiple filter branches to output filtered sample substreams. Each of the multiple filter branches have filter coefficients corresponding to a different phase of the filter response. The filtered sample substreams are stored in a memory and the stored filtered sample substreams are combined to generate the output sample stream.